In message <alpine.NEB.2.00.1206092339570.12986@xxxxxxxxxxxxxxx> TynHau <tide@xxxxxxx> wrote: > On Sat, 9 Jun 2012, Nick Roberts wrote: > > > Date: Sat, 09 Jun 2012 12:13:03 +0100 > > From: Nick Roberts <tigger@xxxxxxxxxxxxxxxxxxxxx> > > Reply-To: armini-support@xxxxxxxxxxxxx > > To: armini-support@xxxxxxxxxxxxx > > Subject: Re: [ARMini-support] ARMini vs Iyonix > > > > In message <alpine.NEB.2.00.1206082216560.4171@xxxxxxxxxxxxxxx> > > TynHau <tide@xxxxxxx> wrote: > > > >> On Fri, 8 Jun 2012, Nick Roberts wrote: > > <snip> > > > > > > I have an Artworks 2 file that is approx 1.7MB (a world map > > > > circa 500BC), with just over 2000 objects. On my Iyonix, this > > > > takes about 2 seconds to render (document loaded, full screen, > > > > and zooming out, so it is strictly a test of the redraw rather > > > > than anything else). > > > > > > > > I copied it across to the ARMini, expecting it to take > > > > signicantly less than that, but it actually takes about 6 > > > > seconds, which is disappointing to say the least[1]. > > > > > > > > Any idea what the holdup is likely to be? Poor memory > > > > bandwidth? Better graphics processing? > > > > <snip> > > > > > What happens if you copy the file to your RAM drive (e.g. > > > Memphis)? > > > > Wny would that make any difference? I thought I'd made it clear > > that I wsn't timing how long it took to load, but how long it took > > to redraw when it was already loaded. > > > > However... > > > > Tried the same experiment, and it took the same time as near as I > > could measure it. > > > > Sorry, misunderstood your question then. Slow rendering has been a > known issue. Hardly a surprise, is it? > > http://www.riscosopen.org/wiki/documentation/show/OMAPVideo I'm not sure how this is relevant. The only things on the "limitations" list that touch on this is accelerated rectangle copy/fill operations, and as there are no rectangles in my test file. I suppose clearing prior to redraw counts, but is that going to be responsible by itself for such a dramatic slowdown? > http://www.riscosopen.org/forum/forums/5/topics/466?page=5 According to the benchmarks in this the table, we have: ARMini (1000MHz) Iyonix (600MHz) Draw Path 329% 90% Draw Fill 364% 66% where the reference is a StrongARM RiscPC. I'm not sure why the Iyonix is so slow compared to a StrongARM, but apart from that the figures would indicate that an ARMini should be 3-5 times faster than an Iyonix, not 1/3rd as fast, which is what my test shows (all the objects in the test document are paths, filled paths, or text (which are in practice more filled paths). -- Nick Roberts tigger @ orpheusinternet.co.uk Hanlon's Razor: Never attribute to malice that which can be adequately explained by stupidity. --- To alter your preferences or leave the group, visit //www.freelists.org/list/armini-support List-related queries to info@xxxxxxxxxxxx